Member of Technical Staff, Machine Learning Engineer

“If you want to do great work, you clearly must work on important problems.” – Richard Hamming, You and Your Research About us Quadrillion is building the research lab in a box. Our product, Qualia, is an agentic research editor: a ground-up replacement for JupyterLab and VSCode that orchestrates agents to design experiments, run them in parallel, and determine what to try next. ML engineers can use this to sweep across models, quants to grind through alt-data datasets, and data scientists to generate beautiful visualizations. The goal is for researchers to spend their time on ideas, not boilerplate. We imagine a future where starting a research lab is as easy as typing in a prompt. We’ve raised a $17M seed round to build this vision, led by Audacious Ventures and Conviction, with participation from ex/ante, Bloomberg Beta, Wes McKinney (founder of Pandas) and many more. Who we’re looking for This is a role for high-agency engineers that latch onto hard problems. We’re looking for “"spiky"” engineers with a willingness to learn. Maybe you’re a full-stack engineer who helped build a startup from zero to one. Perhaps you have deep technical instincts and have strong opinions about how systems should work. Or a researcher who’s always eager to push the latest technology to its limits. We like people with sharp technical taste. Bring strong opinions, weird edge cases you’ve run into, or systems you’ve thought deeply about – tell us your really strongly held opinion about learning rate schedules, or lecture us on some weird Pytorch feature we don’t know! We care less about being right all the time and more about how you reason and iterate. Note: you’ll be writing Python every day on the job, especially PyTorch and HuggingFace library code. This means that we’re looking for people that are highly skilled in one of those two buckets, or are willing to learn fast to become highly skilled. We’re full-time, in-person in NYC. We pay well (cash and equity) and cover lunch/dinner, health insurance, and generous PTO for full-time employees. Why we want you to join us We’ve moved fast with a tight-knit, hardworking team. We work on making our system better by the hour. In just a couple weeks, we built a functioning Jupyter replacement that we use every day. Building a product to take on Jupyter and make it 10x better is already incredibly hard . Making a product that unblocks the path to unsolved problems is even more so. Here’s a taste of the problems we tackle every day: Building a next-gen Jupyter-compatible notebook optimized for LLMs. Designing kernels that permit distributed experiment orchestration, both local and cloud-based. Training models that can actually do experimentation—collecting data, building benchmarks, and standing up infrastructure for large-scale RL rollouts. Building evals to understand what makes an agent reason like a scientist, not just answer like a student. Developing new ways for humans and agents to coordinate inside real research workflows. If you want to help us reinvent the modern research workflow from the ground up, and you’re drawn by the call of unsolved problems – come work with us at Quadrillion. About us We’re a team of ex-quants, founding engineers, and PhDs. Everyone on the team is our own customer: most of us are researchers (2k+ collective cites) and we’ve all spent many late nights writing horrible boilerplate code and cursing out Jupyter notebooks. We’re incredibly determined and relentless, and we expect to take over the market at short notice. We have an amazing team of advisors and investors. Our investors include Audacious (lead), Conviction (lead), Bloomberg Beta, and ex/ante. We’re backed by Adrien at Streamlit, Barry at Hex, Charlie at Astral, Tristan at dbt labs, as well as leaders like Wes McKinney (creator of Pandas), Aparna at Microsoft (creator of Google Lens), and Charlie Songhurst, among many other folks we greatly respect. #J-18808-Ljbffr Q uadrillion